hypnguyen1209/offensive-claude / skills/threat-model-discipline
threat-model-discipline
VerifiedUse when starting an engagement, before exploitation, or whenever the attack surface changes — build/validate the threat model and detect drift (new unreviewed surface) before advancing
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 2.6 KB |
| scripts/threatmodel_lint.py | script | 5.8 KB |
scripts/threatmodel_lint.py
View on GitHub#!/usr/bin/env python3
"""threatmodel_lint.py - enforce a complete threat model, and detect DRIFT across a re-run.
Two jobs:
1. LINT a threat model (JSON) for the fields a usable model must have - trust boundaries, assets,
entry points, ATT&CK techniques, mitigations - and reject placeholder/empty content.
2. DRIFT: diff an updated model against the reviewed baseline. A NEW entry point / asset / trust
boundary that wasn't in the reviewed model is UNREVIEWED attack surface - the novel value for
long-running engagements. `/engage.gate` runs drift and refuses to advance on un-acknowledged drift.
Threat model JSON shape (materialized from recon-osint; see templates/threat-model/):
{"assets":[...], "entry_points":[...], "trust_boundaries":[...],
"attck":["T1190", ...], "mitigations":[...]}
Pure stdlib. CLI:
threatmodel_lint.py lint model.json # exit 1 if incomplete
threatmodel_lint.py drift baseline.json new.json [--json out] # exit 1 if un-acknowledged drift
"""
from __future__ import annotations
import argparse
import json
import re
import sys
REQUIRED_LISTS = ("assets", "entry_points", "trust_boundaries", "attck", "mitigations")
# surfaces where a newly-appeared item means unreviewed attack surface (drift that must be re-reviewed)
DRIFT_SURFACES = ("entry_points", "assets", "trust_boundaries")
_PLACEHOLDER = re.compile(
r"^\s*(tbd|tba|tbc|todo|fixme|placeholder|none|n/?a|fill[\s_-]?in|xxx+|\[.*\]|<.*>|[.\-?]+)\s*$", re.I)
_ATTCK = re.compile(r"^T\d{4}(\.\d{3})?$")
def _as_items(v):
if isinstance(v, list):
return [str(x) for x in v]
return []
def lint(model: dict) -> list:
"""Return a list of issues; empty == complete."""
issues = []
if not isinstance(model, dict):
return ["threat model must be a JSON object"]
for key in REQUIRED_LISTS:
v = model.get(key)
if not isinstance(v, list) or not v:
issues.append(f"'{key}' is missing or empty")
continue
for item in v:
if not isinstance(item, str) or not item.strip() or _PLACEHOLDER.match(item):
issues.append(f"'{key}' has a placeholder/empty entry: {item!r}")
for tid in _as_items(model.get("attck")):
if not _ATTCK.match(tid.strip()):
issues.append(f"'attck' entry is not a valid technique id (Txxxx): {tid!r}")
return issues
def drift(baseline: dict, updated: dict) -> dict:
"""Diff updated vs baseline on the drift surfaces. Added items on a surface = unreviewed."""
if not isinstance(baseline, dict) or not isinstance(updated, dict):
raise ValueError("both threat models must be JSON objects")
# A drift surface supplied as a non-list (dict/string/number) would make _as_items() return []
# and HIDE every added item -> fail-open. Reject it so a malformed model fails CLOSED (exit 2).
for surface in DRIFT_SURFACES:
for label, m in (("baseline", baseline), ("updated", updated)):
v = m.get(surface)
if v is not None and not isinstance(v, list):
raise ValueError(f"{label} '{surface}' must be a list, got {type(v).__name__}")
# items must be strings: an int 5150 vs str '5150' would str-collide and hide a type-change
if isinstance(v, list) and any(not isinstance(x, str) for x in v):
raise ValueError(f"{label} '{surface}' has a non-string entry - fail closed")
report = {"added": {}, "removed": {}, "has_drift": False}
for surface in DRIFT_SURFACES:
base = set(_as_items(baseline.get(surface)))
new = set(_as_items(updated.get(surface)))
added = sorted(new - base)
removed = sorted(base - new)
if added:
report["added"][surface] = added
if removed:
report["removed"][surface] = removed
# only ADDED surface items are drift that blocks the gate (new unreviewed surface);
# removed items are recorded but don't block (surface shrank).
report["has_drift"] = any(report["added"].values())
report["unreviewed_surface"] = sorted(
f"{s}:{item}" for s, items in report["added"].items() for item in items)
return report
def _load(path: str) -> dict:
with open(path, "r", encoding="utf-8") as fh:
return json.load(fh)
def main(argv=None) -> int:
p = argparse.ArgumentParser(description="Lint a threat model + detect drift across a re-run.")
sub = p.add_subparsers(dest="cmd", required=True)
li = sub.add_parser("lint"); li.add_argument("model")
dr = sub.add_parser("drift"); dr.add_argument("baseline"); dr.add_argument("updated"); dr.add_argument("--json")
args = p.parse_args(argv)
try:
if args.cmd == "lint":
issues = lint(_load(args.model))
if issues:
for i in issues:
print(f" - {i}")
print(f"\nthreat model INCOMPLETE: {len(issues)} issue(s)")
return 1
print("threat model complete: all required fields present")
return 0
if args.cmd == "drift":
rep = drift(_load(args.baseline), _load(args.updated))
if args.json:
with open(args.json, "w", encoding="utf-8") as fh:
json.dump(rep, fh, indent=2)
if rep["has_drift"]:
print("DRIFT - unreviewed attack surface added since baseline:")
for item in rep["unreviewed_surface"]:
print(f" + {item}")
print("\nre-review the threat model before advancing (or acknowledge the new surface).")
return 1
print("no drift on entry_points/assets/trust_boundaries")
return 0
except (OSError, ValueError, json.JSONDecodeError) as exc:
print(f"error: {exc}", file=sys.stderr)
return 2
return 2
if __name__ == "__main__":
raise SystemExit(main())
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| rea-analysis-optimizationmorluto/reaOptimize REA providers and shared analysis workflows for agent context, latency, CPU, and memory while preserving correctness and Evidence. Use for analysis bottlenecks and cache or query design; CI/test-suite optimization belongs to rea-ci-tests-audit. | Stars: 472k | Updated | Claude CodeCodex | Verified | Install |
| rea-ci-tests-auditmorluto/reaAudit REA CI and test suites for correctness, fixture drift, false passes, and measured performance; use for suite or workflow investigations and efficiency changes, rather than ordinary feature testing or a single test failure. | Stars: 472k | Updated | Claude CodeCodex | Verified | Install |
| rea-tool-designmorluto/reaAudit, investigate, design, or change REA investigation tools, CLI/MCP contracts, provider capabilities, and Evidence semantics. | Stars: 472k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-auditDietrichGebert/ponytailQuality audit of the whole repo: bugs, security, real load, missing tests, speed, and what to delete. Most important first. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |